Vitamin K

What it actually measures

Vitamin K's job is activation. It attaches a chemical group to certain proteins, and until that happens those proteins do not work.

The proteins in question fall into two groups, which is where the two forms come in.

K1, phylloquinone, comes from leafy greens and is used mainly by your liver, where it activates clotting factors. That is the function vitamin K was discovered for and the reason its deficiency was historically about bleeding.

K2, menaquinone, comes from fermented foods, notably natto, and from some cheeses and animal fats. It is used more by tissues outside the liver, and it activates two proteins that matter here.

Osteocalcin, which helps bind calcium into the bone matrix. And matrix Gla protein, one of your body's inhibitors of calcium deposition in blood vessels.

Without adequate K2, those proteins remain in an uncarboxylated, inactive state. The calcium still circulates. The traffic direction is simply weaker.

Why a normal clotting test says nothing about K2

Here is the measurement problem that makes this marker unusual.

The traditional way vitamin K status was assessed is through clotting: prothrombin time, or INR. If clotting is normal, K status was considered adequate.

That test reflects liver clotting factors, which get priority for whatever vitamin K is available. So your liver can be perfectly supplied, your clotting entirely normal, and the extra-hepatic proteins like matrix Gla protein still substantially uncarboxylated.

A normal INR tells you your clotting factors have enough. It does not tell you your arteries do.

Better status markers exist and are largely research tools rather than routine tests: undercarboxylated osteocalcin, and dephosphorylated-uncarboxylated matrix Gla protein, both measuring the inactive forms that accumulate when K is short.

A 2016 review in Nutrients examining exactly this concluded there is no single gold-standard biomarker of vitamin K status. The undercarboxylated fractions of three vitamin K-dependent proteins are measurable and respond to supplementation, and they correlate only modestly with each other, which means the right biomarker depends on which outcome you care about. It also noted that menaquinones are generally undetectable in circulation unless large amounts are consumed.

That is an unusually candid state of affairs, and it is worth knowing before paying for a test that promises to settle your vitamin K status.

The practical consequence is that most people cannot easily test this, which shifts the question from measurement to intake.

On outcomes, the evidence has improved recently. A 2026 randomized placebo-controlled trial in JAMA Cardiology enrolled 180 patients with coronary calcification scores between 50 and 400 Agatston units, giving 360 micrograms of menaquinone-7 daily or placebo for two years with CT scanning.

Calcification progressed more slowly on treatment, significant at p equals 0.02 after adjustment, with no significant adverse effects. The authors noted the findings suggest MK-7 may slow calcification in noncalcified plaques, and that the clinical significance in terms of plaque stability remains to be determined.

That is a genuine randomized outcome on an imaging endpoint, held at the confidence its authors gave it.

What moves it

Down: low intake of both forms, which is common for K2 specifically since few Western diets include natto. Antibiotics, since gut bacteria produce some K2. Fat malabsorption, since vitamin K is fat soluble, meaning coeliac disease, inflammatory bowel disease, pancreatic insufficiency and bile flow problems. And warfarin, which works by deliberately interfering with vitamin K.

Up: leafy greens for K1, and for K2 natto by a wide margin, followed by some cheeses, egg yolks and animal fats from pasture-raised animals. Supplements, where MK-7 is the form used in most research because of its longer half-life. Your gut bacteria produce some K2, which is part of why antibiotics matter here.

The honest hedge

I am not a doctor and I am not diagnosing anyone, and there is one genuine safety line here that is not optional.

If you take warfarin, do not start vitamin K without your prescriber's involvement. Warfarin works by interfering with vitamin K, so changing your intake changes how the drug behaves and can shift your INR. A steady known intake is what matters, and any change belongs with whoever manages your anticoagulation.

That is not an abundance of caution, it is the mechanism of the drug.

On the calcification evidence, I want to hold it exactly where its authors did. A randomized trial showing slower progression of a surrogate imaging marker over two years is genuinely encouraging. It is not proven cardiovascular protection, and the authors said so themselves.

The status testing situation is honestly poor for consumers. The markers that actually reflect extra-hepatic vitamin K status are largely research tools, so most people are reasoning from intake rather than from measurement.

That is a limitation worth naming rather than papering over with a test that answers a different question.

So. What to actually do.

If you take warfarin, stop here and speak to your prescriber. That is the one hard line on this page.

Tonight, at no cost. Check whether your vitamin D supplement already contains K2. Combined products are common and people frequently double up without noticing.

Eat both forms. Leafy greens for K1, and for K2 natto if you can manage it, otherwise some cheeses, egg yolks and pasture-raised animal fats.

Do not read a normal INR as adequate K status. It reflects liver clotting factors, which get first call on available vitamin K, and says nothing about matrix Gla protein.

If you supplement, MK-7 is the researched form. It has a longer half-life than MK-4 and it is what the calcification trial used.

Think in years, not weeks. The trial needed two years to show its difference, which is a fair indication of how this works.

Keep the rest of cardiovascular care in place. ApoB, blood pressure, inflammation and blood sugar are not replaced by a vitamin.

You are not just absorbing calcium. You are directing it, and the proteins doing the directing need switching on before they can do anything at all.

Your body is not broken. It is blocked. And sometimes the block is a set of proteins sitting in their inactive form, in a system where a normal clotting test looked like an answer.

Go check whether your D already contains K2.